Development of a Questionnaire to Assess the Dietary Behavior of LowIncome Populations

1. Introduction

Understanding what people eat is essential for designing nutrition interventions, yet collecting reliable dietary data from lowincome groups presents unique challenges. Economic constraints, limited food access, cultural preferences, and varying literacy levels all affect how individuals report their eating habits. A welldesigned questionnaire can capture key aspects of dietary behavior while being feasible for field use and sensitive to participants circumstances.

2. Conceptual Framework

The questionnaire should be grounded in a framework that links three domains:

• Food Availability & Access: Physical and economic access to nutritious foods.
• Food Utilization: Knowledge, cooking skills, and cultural preferences that shape food choice.
• Food Security Status: Frequency of skipping meals, reliance on food assistance, and perceived adequacy.

By covering each domain, the instrument can differentiate between lack of choice and lack of resources, which is crucial for tailoring interventions.

3. Item Generation

3.1 Sources of Items

Items were drawn from four primary sources:

• Existing validated tools (e.g., USDA Food Security Survey Module, Dietary Screener Questionnaire).
• Qualitative interviews with community members to capture locally relevant foods and eating patterns.
• Literature on nutrition poverty and food deserts.
• Expert panel review (nutritionists, sociologists, community health workers).

3.2 Types of Questions

To reduce respondent burden, a mixedformat approach is used:

• Frequency items: In the past 7 days, how many times did you eat fruit? (07).
• Portionsize visuals: Small images of a cup, a handful, or a standard serving.
• Binary items: Yes/No questions about use of food assistance.
• Likert scales: I feel I can afford the foods my family likes (Strongly disagree Strongly agree).

4. Draft Questionnaire Structure

1. Screening & Consent brief intro, eligibility, and oral consent.
2. Demographics age, gender, household size, income bracket, education.
3. Food Security 6item short form of USDA module.
4. Food Access distance to nearest grocery store, transportation mode, price perception.
5. Dietary Intake frequency of key food groups (fruits, vegetables, whole grains, sugary drinks, processed meats). Uses a 7day recall.
6. Cooking & Preparation availability of cooking facilities, time spent cooking, use of recipes.
7. Nutrition Knowledge & Attitudes statements about health benefits of foods with true/false options.
8. Barriers & Facilitators openended question: What makes it hard or easy for your household to eat healthily?

5. PreTesting and Cognitive Interviewing

Before largescale administration, the draft was tested with 20 participants from two lowincome neighborhoods. The process included:

• Reading each question aloud and asking respondents to paraphrase the meaning.
• Observing where participants hesitated or asked for clarification.
• Recording suggestions for simpler wording or alternative response options.

Findings led to the removal of jargon (nutrient density), replacement of per week with in the last 7 days, and the addition of a visual portion guide for staple foods.

6. Reliability and Validity Assessment

6.1 Pilot Study

A pilot of 150 households was conducted. Testretest reliability was evaluated by readministering the questionnaire after two weeks to a subsample (n=50). Intraclass correlation coefficients (ICCs) for frequency items ranged from 0.71 to 0.85, indicating good stability.

6.2 Construct Validity

Factor analysis revealed three clear factors aligning with the conceptual framework (Food Access, Food Utilization, Food Security). Items loaded strongly (>0.6) on their expected factor, supporting construct validity.

6.3 Criterion Validity

For a subsample (n=30), 24hour dietary recalls were collected as a gold standard. Correlations between questionnairebased fruit intake frequency and recallderived servings were r = 0.58 (p < .001), demonstrating acceptable criterion validity for brief screeners.

7. Administration Considerations

To ensure high response rates and data quality:

• Mode: Trained interviewers conduct facetoface interviews in community centers or homes.
• Language: Translate into the predominant local languages and backtranslate to confirm accuracy.
• Literacy: Use visual aids and read questions aloud; avoid selfadministered forms where literacy is low.
• Incentives: Small grocery vouchers (e.g., $5) increase participation without influencing responses.

8. Ethical and Cultural Sensitivity

Lowincome participants may feel stigma when discussing food insecurity. Researchers should:

• Emphasize confidentiality and voluntary participation.
• Use neutral wording (e.g., sometimes you may not have enough food) rather than judgmental language.
• Engage community leaders in the design process to respect cultural food practices.

9. Limitations

Even with rigorous development, the questionnaire has constraints:

• Recall bias participants may underreport unhealthy foods.
• Crosssectional design cannot capture seasonal variations in food availability.
• Limited to selfreported data does not directly assess nutritional status.

10. Future Directions

Further work can expand the tools applicability:

• Adapting for mobilephone administration with audio prompts.
• Linking questionnaire responses with objective biomarkers (e.g., serum carotenoids).
• Testing the instrument in diverse geographic settings (rural vs. urban).

Longitudinal use will help monitor the impact of policy changes, such as subsidies for fresh produce or expansions of SNAP benefits.

11. Conclusion

The questionnaire presented here offers a concise, culturally sensitive, and psychometrically sound method for assessing dietary behavior among lowincome populations. By integrating food security, access, and utilization domains, it provides a comprehensive picture that can inform targeted nutrition interventions, guide publicpolicy decisions, and ultimately contribute to reducing dietrelated health disparities.

For researchers interested in using the instrument, a full copy of the questionnaire, scoring guidelines, and dataanalysis scripts are available upon request from the corresponding author.